Antenna structure parameter optimization method based on image translation

Abstract

The invention discloses an antenna structure parameter optimization method based on image translation, and the method comprises the following steps: S1, selecting a to-be-optimized structure parameter according to a basic structure of a target antenna, and determining a target electromagnetic response performance parameter; s2, performing sample generation on the selected structure parameter vector through an LHS algorithm to obtain electromagnetic response data corresponding to the structure parameter; s3, performing image data conversion on the structure parameters and the electromagnetic response data to form a data set, and training an image translation network through the data set; s4, performing image translation on the target electromagnetic response image by using the trained image translation network to obtain a corresponding structure parameter image; and S5, performing parameter extraction on the structure parameter image to obtain structure parameters of the target antenna, and completing antenna design. The method solves the problems that in the prior art, the requirement for priori knowledge of engineers is high, calculation cost and time cost are high, accuracy is insufficient, and generalization of different antenna design tasks is poor.

Description

Technical Field

[0001] The invention relates to the field of mold technology, and in particular to an antenna structure parameter optimization method based on image translation. Background Art

[0002] Antennas are the physical basis for transmitting and receiving wireless signals, and play a core role in improving communication quality and efficiency. Advances in antenna technology have directly promoted the development of wireless communication systems. Traditional antenna design is based on engineers' prior knowledge and a large number of electromagnetic simulations, which will result in a lot of computational and time costs, especially for complex antennas such as multi-frequency antennas and phased array antennas. These costs will be further magnified. Therefore, it is crucial to design an antenna optimization method that is fast, accurate, and less dependent on simulation.

[0003] With the rise of deep learning, many antenna optimization methods based on deep learning have emerged in recent years. For example, Dhaliwal et al. proposed a hybrid algorithm of artificial neural network (ANN) based on particle swarm optimization (PSO) and used it to optimize the antenna size in the 5.8GHz WLAN band. Zhang et al. proposed a deep Gaussian process (DGP) combined with the PSO algorithm to optimize the S11 curve of a multi-band antenna. Liu et al. proposed an ANN-GA hybrid algorithm to optimize the S21 parameters of multiple-input multiple-output (MIMO) antennas to achieve the goal of reducing coupling. Dong et al. used radial basis function neural network (RBFNN) as a proxy model to replace full-wave simulation, and combined with an improved PSO algorithm to optimize the impedance matching effect and size of the antenna. Farzad Mir et al. used deep neural network (DNN) as a proxy model, combined with Thompson sampling efficient multi-objective optimization (TSEMO), to optimize the S11 curve and gain of the proposed microstrip antenna. The above methods use ANN, DGP, etc. as proxy models to replace full-wave simulation, and define optimization problems (such as S11 curve, antenna size, etc.) and combine PSO, GA and other optimization algorithms to achieve efficient iterative search to find the optimal solution. Although this forward design method uses a proxy model to quickly obtain simulation results, it still requires multiple iterations to achieve the optimization goal. At the same time, the prediction results obtained by the proxy model may have errors, and this error will be amplified during multiple iterations and is not conducive to finding the optimal solution. In addition, when there are multiple design tasks, the forward design method needs to search and iterate for different tasks separately, and the time and computational cost will increase exponentially.

[0004] Compared with forward design, reverse design directly obtains the corresponding antenna structure parameters by taking the target electromagnetic response as input, eliminating the iterative process of the optimization algorithm and is a more efficient method. Sharma et al. sampled the S11 curve of a dual-band T-type antenna and expressed the sum of the absolute values ​​of the sampling points as FOM. They used ANN and other algorithms to predict FOM to achieve the purpose of optimizing impedance matching and bandwidth. et al. compressed the S11 curve spectrum of a dual-frequency antenna into four point values ​​as data to train ANN and optimize the four parameters of the proposed dual-frequency antenna. Singh et al. constructed a branch output convolutional neural network architecture as an inverse model, trained the model with 10,000 sets of data, and predicted the geometric parameters of the proposed polygonal patch antenna. Kabir et al. divided the data set into several groups and trained multiple networks, combining them into an inverse model for filter performance optimization. Although these studies have achieved the reverse design of antennas, there are generally problems such as inaccurate prediction of target electromagnetic response, excessive amount of training data required for the model, and poor generalization. Therefore, it is urgent to propose an efficient, accurate, low-data-dependence and strong generalization method for antenna inverse design. Summary of the invention

[0005] The purpose of the present invention is to provide an antenna structure parameter optimization method based on image translation, which solves the problems in the prior art such as high requirements on engineers' prior knowledge, high calculation cost and time cost, insufficient accuracy, and poor generalization for different antenna design tasks.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] An antenna structure parameter optimization method based on image translation comprises the following steps:

[0008] S1. According to the basic structure of the target antenna, select the structural parameters that need to be optimized and determine the target electromagnetic response performance parameters and the number of training samples.

[0009] S2. Samples of the selected structural parameter vector are generated through the Latin hypercube sampling algorithm LHS, and the electromagnetic response data corresponding to the structural parameters are obtained through the Python-HFSS joint simulation script.

[0010] S3. Convert the structural parameters and electromagnetic response data into image data to form a data set, and use the data set to train the image translation network.

[0011] S4, using the trained image translation network to perform image translation on the target electromagnetic response image to obtain a corresponding structural parameter image;

[0012] S5. Extract parameters from the structural parameter image to obtain the structural parameters of the target antenna, and complete the antenna design.

[0013] Preferably, the structural parameters in S1 include the length, width and distance to the edge of the substrate of each component in the antenna; the electromagnetic response performance parameters include the S11 curve and the radiation pattern.

[0014] Preferably, the sample generation of the structure parameter vector includes the following steps:

[0015] First, the sample size and parameter dimension are determined, and the range of each parameter dimension is divided into N equal probability intervals.

[0016] Secondly, a sample point is randomly selected in each equal probability interval of each parameter dimension and randomized.

[0017] Finally, the sample points of each parameter dimension are combined to form N samples.

[0018] Preferably, in S3, the image translation network is an improved pixpix image translation network, including a generator and a discriminator.

[0019] The generator uses U-Net as the backbone network. In order to strengthen the feature extraction and processing of multiple image channel superpositions, the SE attention mechanism is added between the upsampling layer, downsampling layer and jump layer connections in U-Net. At the same time, the CondConv dynamic convolution enhanced translation network is used to replace the ordinary convolution in the upsampling layer and downsampling layer to further improve the feature extraction capability.

[0020] The SE attention mechanism is also added between the downsampling layers of the discriminator, and the CondConv dynamic convolution enhanced translation network is used to replace the ordinary convolution.

[0021] Preferably, the CondConv dynamic convolution enhanced translation network calculates the weights of different convolution kernels according to the input features before ordinary convolution, and dynamically calculates the convolution kernels for different inputs, so that the network can adjust its convolution operation according to the characteristics of each data sample, so as to better process the data characteristics of the superimposed channel, and finally obtain the dynamic convolution kernel by weighted summation of each convolution kernel.

[0022] The calculation expression is:

[0023] Output(x)=σ((α1W1+…+α n W n )*x);

[0024] Where σ is the activation function; W1,...,W n is the convolution kernel; α1,...,α n is the weight corresponding to each convolution kernel; x is the input feature map.

[0025] Preferably, the loss function of the generator is composed of a generative adversarial loss function and a pixel loss function. The generative adversarial loss is used to guide the model to converge in the direction of accurately fitting the real electromagnetic response. The pixel loss is used to adjust the consistency of the generator output image and the real value in the surface semantics (such as image color, basic shape). The adversarial loss function is:

[0026]

[0027] The pixel loss function is:

[0028]

[0029] In the formula, Image GT and Image G The real images and the images generated by the generator respectively; and are the generative adversarial losses for the conditional information path and the noise path, respectively; Pixel loss for applying smoothL1; is the final generator loss; G and D are the generator and discriminator respectively; Image and Noise are the input conditional information image and random noise image respectively; is the final generator loss; is the expected value of the image under the distribution of training set data; is the expected value of the random noise distribution.

[0030] Preferably, S5 further comprises the following steps:

[0031] The performance of the trained image translation network is measured by calculating the relative error between the electromagnetic response corresponding to the structural parameters output by the image translation network and the real electromagnetic response:

[0032]

[0033] Where, misfit1 and misfit2 are the relative error values ​​of the S11 curve and the radiation pattern, respectively. The lower the error value, the higher the prediction accuracy of the image translation network; n is the number of test samples; They are the S11 curve and radiation pattern predicted and fitted by the image translation network; They are the label values ​​of the S11 curve and radiation pattern, respectively.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] The present invention can translate the input electromagnetic response image into the corresponding structural parameter image within a few seconds through the improved image translation network, eliminating the iterative search process. In addition, the amount of training data required for the existing deep learning algorithm training is greatly reduced while maintaining high prediction accuracy. The time cost and computational cost required for the entire optimization process are only in the data collection and model training stages. There is no need for an iterative search process for multiple design tasks for the same antenna. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of the structure of the multi-frequency dipole antenna according to an embodiment of the present invention; (a) is a schematic diagram of the front structure; (b) is a schematic diagram of the back structure;

[0037] Figure 2 This is a schematic diagram of the optimization method of the present invention;

[0038] Figure 3 A schematic diagram of converting one-dimensional vector data into image data according to an embodiment of the present invention;

[0039] Figure 4 Schematic diagram of the structure of the image translation network according to an embodiment of the present invention; (a) is a schematic diagram of the structure of the generator; (b) is a schematic diagram of the structure of the discriminator; (c) is a schematic diagram of the detailed structure of the upsampling layer and the downsampling layer in the generator and the discriminator;

[0040] Figure 5 This is a schematic diagram of the SE attention mechanism principle of an embodiment of the present invention;

[0041] Figure 6 This is a schematic diagram of the CondConv dynamic convolution principle of an embodiment of the present invention;

[0042] Figure 7 A schematic diagram of the training process of an image translation network according to an embodiment of the present invention;

[0043] Figure 8 Schematic diagram of some test results in the test set of the embodiment of the present invention; wherein (a) is a comparison of the results of test example 1 in Table 3; (b) is a comparison of the results of test example 2 in Table 3; and (c) is a comparison of the results of test example 3 in Table 3. DETAILED DESCRIPTION

[0044] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0045] See also Figure 1-Figure 8, an antenna structure parameter optimization method based on image translation, generates random structure parameter samples, collects data automatically through joint simulation scripts, and constructs a data set after data conversion of the original data. The image translation network is trained, and the trained model is used to quickly implement various antenna design tasks. Finally, the model is tested and evaluated.

[0046] The specific steps include:

[0047] S1. According to the basic structure of the target antenna, select the structural parameters that need to be optimized and determine the target electromagnetic response performance parameters; optimize the impedance matching and radiation pattern to obtain good resonance and omnidirectional radiation at the target frequency.

[0048] Specifically, the initial structural model of the target antenna is preliminarily constructed in the simulation software HFSS, such as Figure 1 As shown, the target antenna uses Rogers RT Duroid 5880 with a dielectric constant Er=2.2 and a loss tangent tanδ=0.0009 as a substrate material, and the size of the substrate is 100 mm*55 mm*0.787 mm.

[0049] The design index is: the return loss value at the three frequency points of 0.9GHz, 1.8GHz and 2.4GHz is lower than -10dB to achieve good resonance. And the non-circularity of the XoZ plane radiation pattern at these three frequency points is less than 3dB (the difference between the maximum and minimum values ​​of the pattern curve) to achieve omnidirectional radiation characteristics. There are 12 main parameters that determine the performance of the antenna in this embodiment, as shown in Table 1.

[0050] Table 1 Target antenna performance parameters

[0051]

[0052]

[0053] S2. Samples of the selected structural parameter vector are generated through the Latin hypercube sampling algorithm LHS, and the electromagnetic response data corresponding to the structural parameters are obtained through the Python-HFSS joint simulation script.

[0054] Use vector P to represent the structural parameters, that is, P = [L1, L2, L3, L4, L5, L6, L7, W1, W2, W3, W4] T. Use the Latin Hypercube Sampling algorithm (LHS) to generate 800 sets of structural parameters, then use Python to automatically call the HFSS interface, and input the structural parameters generated by LHS into HFSS for simulation to obtain the corresponding electromagnetic response data. P, S, and R represent the structural parameters, S11 curve, and radiation pattern, respectively. Each structural parameter P and its corresponding parameter S and radiation pattern R are taken as a data sample to form a data set. Among them, 600 groups are used for training the model, 150 groups are used for verification, and 50 groups are used to test the model performance.

[0055] The Latin Hypercube Sampling algorithm (LHS) comprises the following steps:

[0056] Determine the number of samples N and the parameter dimension D, where the parameter dimension D is the number of elements in the structural parameter vector P. For each parameter dimension D, divide its range into N equally probable intervals, ensuring that each equally probable interval is evenly considered during the sampling process. In each equally probable interval of each parameter dimension D, randomly select a sample point. Then, randomize the sample points in each parameter dimension D. Finally, combine the sample points of each parameter dimension D to form N samples.

[0057] S3, converting the structural parameters and electromagnetic response data into image data to form a data set, and training an image translation network through the data set;

[0058] In order to convert the one-dimensional vector structure parameter P, S11 curve S and radiation pattern R into corresponding image data while keeping the semantics of the data unchanged, the following three transformations are performed respectively:

[0059] For the structural parameter P, the blank image with a size of 256 pixels * 256 pixels is divided into 12 rectangular areas of the same size at equal intervals in the vertical direction, and filled with black pixels according to the proportional coefficient η. The relationship between the proportional coefficient η and the corresponding parameter in the structural parameter P can be expressed linearly:

[0060]

[0061] Therefore, the proportion of black pixels in the 12 rectangular areas can directly correspond to the 12 parameters in the structural parameter P. The complete image after filling is the structural parameter image P′:

[0062] P′={p′1,p′2…p′ 12};

[0063] in,

[0064] p′ i =η i ×s,i=1,2…12;

[0065] Where ηi is the proportional coefficient of the i-th parameter; p i is the i-th parameter in the structural parameter P, p imin is p in Table 1 i The left edge of the range, p imax is p in Table 1 i The right boundary of the range; s is the area of ​​12 rectangular regions of equal size.

[0066] For the S11 curve S: first draw its curve on a blank image of size 256 pixels * 256 pixels, and fix the horizontal and vertical coordinate ranges of the image so that each pixel value can directly correspond to the data features expressing the electromagnetic response curve. In order to increase the pixel ratio in the image after the one-dimensional vector conversion. Fold the S11 curve symmetrically and fill the area surrounded by the two curves with black pixels. Then rotate the image 90° in the opposite direction to obtain S′, and adjust the image size to 256 pixels * 256 pixels. This image transformation enriches the image features while keeping the semantic information unchanged.

[0067] For the radiation pattern R: the blank image of size 256 pixels * 256 pixels is divided into three areas of equal size vertically, and the radiation pattern curves of 0.9GHz, 1.8GHz, and 2.4GHz are drawn respectively. Unlike the S11 curve, the parameter values ​​of the radiation pattern R are both positive and negative. For the convenience of representation, the area surrounded by the radiation pattern curve and the horizontal straight line at the 0dB value is filled to obtain R'. The three areas are filled with pixel values ​​of RGB (255, 0, 0), RGB (0, 255, 0), and RGB (0, 0, 255) respectively, so that the feature values ​​in the three areas of the RGB three channels of R' do not interfere with each other, so as to achieve the best feature representation effect.

[0068] The image translation network adopts an improved pixpix image translation network, including a generator and a discriminator;

[0069] like Figure 4 As shown in the figure, the generator uses U-Net as the backbone network, and adds the SE attention mechanism between the upsampling layer, downsampling layer and jump layer connections in U-Net. At the same time, the CondConv dynamic convolution enhanced translation network is used to replace the ordinary convolution in the upsampling layer and downsampling layer to further improve the feature extraction capability.

[0070] The SE attention mechanism is also added between the downsampling layers of the discriminator, and the CondConv dynamic convolution enhanced translation network is used to replace the ordinary convolution.

[0071] like Figure 5As shown, the SE attention mechanism includes the Squeeze stage and the Excitation stage. The process of generating an attention feature map through the SE attention mechanism is as follows:

[0072] In the Squeeze stage, after performing a global average pooling operation on the input feature map U of size H×W×C, the spatial information (width and height) of each channel is compressed into a single global feature, resulting in a feature map of size 1×1×C.

[0073] In the Excitation stage, the 1×1×C feature map obtained in the above steps is first reduced in dimension (down-sampled) and then increased in dimension (upsampling). After the dimensionality reduction and dimensionality increase processing, the obtained feature map is normalized to generate the final channel attention weight.

[0074] The attention weights are multiplied element-by-element with the original feature map to generate the feature map X with attention. The SE attention mechanism distinguishes the importance of different features for model learning through attention, helping the model learn important features.

[0075] like Figure 6 As shown in the figure, the CondConv dynamic convolution enhanced translation network calculates the weights of different convolution kernels according to the input features before the ordinary convolution, and finally obtains the dynamic convolution kernel by weighted summation of each convolution kernel; the calculation expression is:

[0076] Output(x)=σ((α1W1+…+α n W n )*x);

[0077] Where σ is the activation function; W1,...,W n is the convolution kernel; α1,...,α n is the weight corresponding to each convolution kernel; x is the input feature map.

[0078] By dynamically calculating the convolution kernel for different inputs, the network can adjust its convolution operation according to the characteristics of each data sample, thereby better processing the data characteristics of the superimposed channels.

[0079] The loss function of the generator is composed of a generative adversarial loss function and a pixel loss function. The generative adversarial loss is used to guide the model to converge in the direction of accurately fitting the real electromagnetic response. The pixel loss is used to adjust the consistency of the generator output image and the real value in the surface semantics (such as image color, basic shape). The adversarial loss function is:

[0080]

[0081] The pixel loss function is:

[0082]

[0083] In the formula, Image GT and Image G The real images and the images generated by the generator respectively; and are the generative adversarial losses for the conditional information path and the noise path, respectively; Pixel loss for applying smoothL1; is the final generator loss; G and D are the generator and discriminator respectively; Image and Noise are the input conditional information image and random noise image respectively; is the final generator loss; is the expected value of the image under the distribution of training set data; is the expected value of the random noise distribution.

[0084] like Figure 7 The figure shows a schematic diagram of the training process of the image translation model. In this embodiment, the conditional information (i.e., electromagnetic response image) and random noise are fed into the network through two paths respectively, and two predicted outputs (conditional output and noise output) are obtained through the generator. Then, the discriminator judges the true or false data pairs composed of the two outputs and the label input and the corresponding label data pairs, and outputs two discriminant values ​​for calculating the generation of adversarial loss back propagation to update the network parameters of the generator and the discriminator.

[0085] In the conditional information path, the generator is able to learn the intrinsic mapping relationship between the antenna electromagnetic response and the structural parameters. In the noise path, a noise bottleneck module is added, which consists of a downsampling layer and a SE attention layer. The noise bottleneck module limits the flow of noise information so that the generator in the improved pixpix image translation network does not directly extract data features from the noise, but learns how to map the noise to the target distribution. This noise-driven process helps the model learn the entire distribution of the target domain. The dual input path structure allows the model to simultaneously learn features from real images and the randomness provided by noise, so as to achieve better learning of the data distribution of the output domain under a limited data range.

[0086] S4. Use the trained image translation network to translate the target electromagnetic response image to obtain the corresponding structural parameter image; the dual input path structure enhances the network's ability to learn the randomness of noise, helping the network to achieve better training results with a small amount of data, further reducing simulation costs. Use the SE attention mechanism and CondConv dynamic convolution to enhance the translation network's ability to extract features from multi-channel input data, ensuring that the model accurately learns important information in the antenna's electromagnetic response.

[0087] By inputting the target electromagnetic response image into a well-trained image translation network, the corresponding structural parameter image can be obtained within a few seconds, and the optimization can be completed directly without repeated optimization search.

[0088] S5. Extract parameters from the structural parameter image to obtain the structural parameters of the target antenna, and complete the antenna design.

[0089] After the target electromagnetic response image is input into the image translation network to obtain the output image, the output structural parameter image is subjected to parameter extraction to obtain the structural parameters. This parameter extraction process is the inverse of the process of converting the structural parameter vector P into the image P′ in the S3 data conversion.

[0090] After extracting the structural parameters of the target antenna, error calculation can be performed to verify whether the parameters meet the requirements. The performance of the trained image translation network is measured by calculating the relative error between the electromagnetic response corresponding to the structural parameters output by the image translation network and the actual electromagnetic response:

[0091]

[0092] Where, misfit1 and misfit2 are the relative error values ​​of the S11 curve and the radiation pattern respectively; n is the number of test samples; They are the S11 curve and radiation pattern predicted and fitted by the image translation network; They are the label values ​​of the S11 curve and radiation pattern, respectively.

[0093] In this embodiment, the 50 groups of data in the test set are subjected to image translation to extract the structural parameters predicted by the model, and the corresponding electromagnetic responses are input into the simulation to calculate the relative errors with the true values.

[0094] As shown in Table 2, Er1, Er2, and Er3 represent the relative errors between the radiation patterns of the model-predicted antenna structure at 0.9 GHz, 1.8 GHz, and 2.4 GHz and the true values, respectively, while Es represents the relative errors between the S11 curves corresponding to the structural parameters output by the model and the true values. The relative error values ​​are all less than 0.1, indicating that the image translation network can fit the input target electromagnetic response well.

[0095] Table 2 Relative error table of test set

[0096] Es Er1 Er2 Er3 0.096 0.051 0.082 0.079

[0097] In order to show the results more intuitively, some comparison results are selected in the test set. Figure 8As shown, the S11 curve and radiation pattern curve corresponding to the antenna structure predicted by the image translation network of this embodiment and the S11 curve and radiation pattern curve corresponding to the actual antenna structure have almost no deviation, indicating that the proposed image translation network has good electromagnetic response fitting performance. The following Table 3 compares its structural parameters with the actual values:

[0098] Table 3 Comparison of the actual values ​​of structural parameters of some examples in the test set and the predicted structural parameter values ​​of the present invention

[0099] <![CDATA[L1]]> <![CDATA[L2]]> <![CDATA[L3]]> <![CDATA[L4]]> <![CDATA[L5]]> <![CDATA[L6 <!-- 8 -->]]> Test Example 1 38.92 / 36.86 32.36 / 32.82 26.66 / 26.09 46.26 / 44.12 11.38 / 11.44 46.99 / 45.69 Test Example 2 36.49 / 35.75 31.52 / 31.74 26.09 / 25.75 44.89 / 43.07 12.24 / 11.83 44.54 / 46.49 Test Example 3 36.67 / 36.78 32.98 / 33.23 24.09 / 24.15 43.04 / 41.96 12.26 / 12.04 47.77 / 47.43 <![CDATA[L7]]> <![CDATA[L G ]]> <![CDATA[W1]]> <![CDATA[W2]]> <![CDATA[W3]]> <![CDATA[W4]]> Test Example 1 78.47 / 77.80 34.54 / 33.62 2.71 / 2.54 5.74 / 5.92 2.05 / 2.09 0.83 / 0.96 Test Example 2 78.43 / 78.02 39.41 / 38.07 2.06 / 2.14 5.29 / 5.68 2.08 / 1.79 0.98 / 1.01 Test Example 3 76.70 / 76.54 34.72 / 34.81 2.45 / 2.30 5.81 / 5.59 2.29 / 2.06 0.99 / 0.91

[0100] As shown in Table 3, the structural parameters predicted by the model are not exactly the same as the true values, but the electromagnetic responses are very similar. The image translation network has learned the potential distribution of the data well, rather than just the surface relationship of the training data, which reflects good data generalization.

Claims

1. An antenna structure parameter optimization method based on image translation, characterized in that: The following steps are involved: S1. According to the basic structure of the target antenna, select the structural parameters that need to be optimized and determine the target electromagnetic response performance parameters; S2, generate samples of the selected structural parameter vectors through the Latin hypercube sampling algorithm LHS, and obtain the electromagnetic response data corresponding to the structural parameters through the Python-HFSS joint simulation script; S3, converting the structural parameters and electromagnetic response data into image data to form a data set, and training an image translation network through the data set; S4, using the trained image translation network to perform image translation on the target electromagnetic response image to obtain a corresponding structural parameter image; S5. Extract parameters from the structural parameter image to obtain the structural parameters of the target antenna, and complete the antenna design.

2. The method for optimizing antenna structure parameters based on image translation according to claim 1, characterized in that: The structural parameters in S1 include the length, width and distance to the edge of the substrate of each component in the antenna; the electromagnetic response performance parameters include the S11 curve and the radiation pattern.

3. The antenna structure parameter optimization method based on image translation according to claim 1, characterized in that: The sample generation of the structure parameter vector includes the following steps: First, determine the sample size and parameter dimension, and divide the range of each parameter dimension into N equal probability intervals; Secondly, a sample point is randomly selected in each equal probability interval of each parameter dimension and randomized; Finally, the sample points of each parameter dimension are combined to form N samples.

4. The method for optimizing antenna structure parameters based on image translation according to claim 1, characterized in that: In S3, the image translation network is an improved pixpix image translation network, including a generator and a discriminator; The generator uses U-Net as the backbone network, adds the SE attention mechanism to the upsampling layer, downsampling layer and jump layer connections in U-Net, and uses the CondConv dynamic convolution enhanced translation network to replace the ordinary convolution in the upsampling layer and downsampling layer; The SE attention mechanism is also added between the downsampling layers of the discriminator, and the CondConv dynamic convolution enhanced translation network is used to replace the ordinary convolution.

5. The method for optimizing antenna structure parameters based on image translation according to claim 4, characterized in that: The CondConv dynamic convolution enhanced translation network calculates the weights of different convolution kernels according to the input features before the ordinary convolution, and finally obtains the dynamic convolution kernel by weighted summation of each convolution kernel; the calculation expression is: Output(x)=σ((α1W1+…+α n W n )*x); Where σ is the activation function; W1,...,W n is the convolution kernel; α1,...,α n is the weight corresponding to each convolution kernel; x is the input feature map.

6. The method for optimizing antenna structure parameters based on image translation according to claim 4, characterized in that: The loss function of the generator is composed of a generation adversarial loss function and a pixel loss function, and the adversarial loss function is: The pixel loss function is: In the formula, Image GT and Image G The real images and the images generated by the generator respectively; and are the generative adversarial losses for the conditional information path and the noise path, respectively; Pixel loss for applying smoothL1; is the final generator loss; G and D are the generator and discriminator respectively; Image and Noise are the input conditional information image and random noise image respectively; is the final generator loss; is the expected value of the image under the distribution of training set data; is the expected value of the random noise distribution.

7. The method for optimizing antenna structure parameters based on image translation according to claim 1, characterized in that: S5 also includes the following steps: The performance of the trained image translation network is measured by calculating the relative error between the electromagnetic response corresponding to the structural parameters output by the image translation network and the real electromagnetic response: Where, misfit1 and misfit2 are the relative error values ​​of the S11 curve and the radiation pattern respectively; n is the number of test samples; They are the S11 curve and radiation pattern predicted and fitted by the image translation network; They are the label values ​​of the S11 curve and radiation pattern, respectively.

Images